When several meals arrive at once
A web-building spider usually has time to detect, wrap and immobilize one struggling insect before settling down to feed. But what happens when several flies land almost at once and each one may escape while the spider is dealing with the others?
Researchers presented 121 long-bodied cellar spiders, Pholcus phalangioides, with one of three laboratory challenges: a single fly, three flies placed on the web simultaneously, or three flies introduced with short delays. The spiders completed 541 trials in total. The multi-prey treatments were deliberately difficult, and the simultaneous three-fly challenge produced more failures than either the one-fly control or the delayed treatment.
Three solutions to the same problem
The researchers recorded 43 successful solutions to the multi-prey problems. The spiders handled the extra flies in three visibly different ways:
- Leave them in place: 11 times, a spider secured each fly separately in the web and moved away with one to feed.
- Gather them: 26 times, a spider brought individually wrapped flies closer together before feeding.
- Bundle them: 6 times, a spider wrapped two flies together in a single silk package.
Gathering was the most common solution and led to the highest proportion of flies being captured. The spiders also used suspension lines, their chelicerae and their hind legs to store or carry prey while working on another fly.
No single routine took over
Individual spiders did not reliably repeat the same solution when they succeeded again. They also did not become consistently faster across successful trials, and the researchers found no decline in failures after a spider had solved the problem. How well a solution worked on one trial did not predict which method the spider used next.
That pattern matters because different kinds of problem-solving make different predictions. A strongly innate routine should vary relatively little. A solution learned through trial and error should become faster or more reliable with practice. Instead, the spiders varied their approach without showing a clear learning curve.
Does that count as insight?
The observed pattern matched the researchers' predictions for insight more closely than their predictions for an innate, previously learned or newly learned solution. In this context, insight does not mean human-like reasoning. It means arriving at a suitable solution without gradually acquiring it through repeated trial and error.
One possibility is that a cellar spider can use information about its web and the positions of trapped prey to construct a solution each time the problem appears. If so, the result would add to growing evidence that sophisticated-looking flexibility does not require a large vertebrate brain.
Why the conclusion remains tentative
The experiment did not directly reveal what the spiders perceived or represented mentally. Web geometry, fly positions, spider orientation and previous experience could not be made perfectly identical between trials. Those differences may have influenced which response was practical at a particular moment.
Repeated successes were also limited and often widely separated in time: the average interval was about nine days, with some reaching 40 days. That design may have made gradual learning difficult to detect. Some flies escaped, and not every successful spider could be observed solving the challenge several times.
The study therefore provides evidence consistent with insight, not proof that cellar spiders consciously planned their next move. Its strongest result is simpler and still fascinating: faced with the same general prey-capture problem, these spiders were not locked into one inflexible sequence.